Vimeo Locked-Embed Caption Extraction

Problem

A page embeds a Vimeo video using </code>.</p><p>The video is "domain-restricted" — opening the player URL directly returns a privacy error,</p><p>and <code>yt-dlp</code> / browser automation also fail because Vimeo enforces the Referer check.</p><p>But you only need the <strong>transcript</strong>, not the video. Vimeo's auto-generated captions are</p><p>served from a separate signed URL that the player HTML embeds in a JSON config — and that</p><p>HTML is fetched whenever the Referer matches the allowed domain. So you can pull the</p><p>captions without ever playing the video.</p><h2>Trigger Conditions</h2><ul><li>Page contains <code><iframe src="https://player.vimeo.com/video/..."></code> and no transcript text.</li><li>Direct visit to the player URL shows "Sorry / 抱歉 — video cannot be played here due to privacy settings."</li><li>You want a transcript-driven analysis (interview, panel, fireside chat, conference talk).</li></ul><h2>Solution</h2><p>Three commands. Replace <code>{ID}</code>, <code>{HASH}</code>, and <code>{HOST}</code> with values from the embed.</p><pre><code># 1. Fetch the player HTML with the correct Referer and grep out the text_tracks JSON. curl -s -H 'Referer: https://{HOST}/' \ 'https://player.vimeo.com/video/{ID}?h={HASH}' \ | grep -oE '"text_tracks":\[[^]]*\]' </code></pre><p>That returns something like:</p><pre><code>"text_tracks":[{"id":303147651,"lang":"en-x-autogen", "url":"https://captions.vimeo.com/captions/303147651.vtt?expires=...&sig=...", "kind":"subtitles","label":"English (auto-generated)", "provenance":"ai_generated","default":true}] </code></pre><pre><code># 2. Download the VTT (the signed URL works without Referer). curl -s 'https://captions.vimeo.com/captions/{CAP_ID}.vtt?expires=...&sig=...' \ -o /tmp/transcript.vtt # 3. Strip WEBVTT cues/timestamps to get plain text. awk '/-->/{next} /^[0-9]+$/{next} /^WEBVTT/{next} /^$/{next} {print}' \ /tmp/transcript.vtt > /tmp/transcript.txt wc -w /tmp/transcript.txt </code></pre><h2>How to find ID, HASH, HOST</h2><p>From the embed iframe's <code>src</code> attribute on the host page:</p><pre><code>https://player.vimeo.com/video/1195836424?h=a0154d0f4b └────┬────┘ └────┬────┘ ID HASH </code></pre><p><code>HOST</code> is the hostname of the page that embeds the iframe (e.g. <code>www.coatue.com</code>). If you</p><p>don't have the iframe URL, open the host page in playwright/devtools and inspect:</p><pre><code>[...document.querySelectorAll('iframe')].map(f => f.src) </code></pre><h2>Verification</h2><ul><li>VTT file should be > 1 KB and contain <code>WEBVTT</code> header + numbered cues.</li><li>Stripped text file word count should be plausible for the video length (≈ 150 wpm</li></ul><p> for normal speech).</p><h2>Notes</h2><ul><li>The VTT is <strong>auto-generated</strong> (<code>provenance":"ai_generated"</code>). Proper nouns and</li></ul><p> brand-name acronyms are often misheard ("Computer Use" → "CP", "Cherny" → "Cherney").</p><p> Re-read the transcript with that bias in mind, especially for names, product</p><p> codenames, and numbers.</p><ul><li>The signed URL has an <code>expires</code> parameter — typically valid for many days, but if</li></ul><p> it 403s, re-fetch step 1 to get a fresh signature.</p><ul><li>If <code>text_tracks</code> returns <code>[]</code>, the video has no captions enabled. Falling back to</li></ul><p> <code>yt-dlp --write-auto-subs</code> won't help (same Referer block); use Whisper on a screen</p><p> recording instead.</p><ul><li>This works because Vimeo enforces playback restrictions on the <strong>video stream</strong> but</li></ul><p> not on the player's HTML config payload, which leaks the captions URL. This has been</p><p> the behavior for years; if Vimeo ever closes it, the fallback is to drive the embed</p><p> inside playwright with the correct Referer and read <code>player.getTextTracks()</code> via the</p><p> Vimeo Player API.</p><ul><li>For multilingual content, <code>text_tracks</code> is an array — check <code>lang</code> field for other</li></ul><p> available subtitle tracks beyond auto-generated English.</p><h2>Example</h2><p>Coatue × Boris Cherny interview, May 2026:</p><ul><li>Embed page: <code>https://www.coatue.com/blog/video/interview-with-claude-code-creator</code></li><li>Iframe src: <code>https://player.vimeo.com/video/1195836424?h=a0154d0f4b</code></li><li>Direct visit → "由于隐私设置，该视频无法在此处播放"</li><li>One curl with <code>Referer: https://www.coatue.com/</code> → captions URL leaked</li><li>Final transcript: 881 VTT lines / 2309 words / ~12 minutes of dialogue.</li></ul><h2>References</h2><ul><li><a href="https://developer.vimeo.com/player/sdk/reference#texttrack" target="_blank" rel="noopener">Vimeo Player text-tracks docs</a></li></ul><p> (Player API; useful as a fallback when the HTML scrape stops working)</p><ul><li>WebVTT spec: <https://www.w3.org/TR/webvtt1/></li></ul></div> </div> </div> <div id="tab-versions" class="detail-content"> <div class="detail-section"> <h2>版本历史</h2> <p style="margin-bottom:12px;font-size:14px;color:#94a3b8;">共 1 个版本</p> <ul class="version-list"> <li> <div> <span class="version-tag">v1.0.0</span> <span style="font-size:11px;color:#5b6abf;margin-left:8px;background:#eef0ff;padding:1px 8px;border-radius:10px;">当前</span> </div> <div style="font-size:12px;color:#94a3b8;"> 2026-06-01 12:52 </div> </li> </ul> </div> </div> <div id="tab-security" class="detail-content"> <div class="detail-section"> <h2>安全检测</h2> <div class="sec-grid"> <div class="sec-card"> <h4>腾讯云安全 (Keen)</h4> <div class="sec-status sec-queued"> 队列中 </div> </div> <div class="sec-card"> <h4>腾讯云安全 (Sanbu)</h4> <div class="sec-status sec-queued"> 队列中 </div> </div> </div> </div> </div>  <div style="margin-top:24px;"> <h2 style="font-size:18px;font-weight:600;margin-bottom:16px;">🔗 相关推荐</h2> <div class="rec-grid"> <div class="rec-card"> <span class="badge-cat" style="margin-bottom:8px;display:inline-block;"></span> <h3><a href="/s/fn-ime-voice-switch">Fn Ime Voice Switch</a></h3> <div class="rec-owner">heavenchenggong</div> <div class="rec-desc">macOS 上用 Hammerspoon 实现按住 Fn 切到豆包语音输入、松开切回默认输入法的自动化方案。用户说「设置 Fn 切换输入法」/「Fn 切到豆包语音」/「按 Fn 切豆包」/「macOS 输入法 Fn 自动化」/「Hammer</div> <div class="rec-stats"> <span style="color:#f39c12;">★ 0</span> <span style="color:#5b6abf;">📥 226</span> </div> </div> <div class="rec-card"> <span class="badge-cat" style="margin-bottom:8px;display:inline-block;"></span> <h3><a href="/s/evernote-note">Evernote Note</a></h3> <div class="rec-owner">heavenchenggong</div> <div class="rec-desc">印象笔记 API 技能，管理用户的印象笔记。支持搜索、浏览笔记本，获取、新建、追加笔记内容。v2.0 新增：通过官方 RESTful API 调用网页剪藏（Web Clipper），抓取任意 URL 页面正文并存入笔记。用户提到印象笔记、E</div> <div class="rec-stats"> <span style="color:#f39c12;">★ 3</span> <span style="color:#5b6abf;">📥 1,180</span> </div> </div> <div class="rec-card"> <span class="badge-cat" style="margin-bottom:8px;display:inline-block;"></span> <h3><a href="/s/personal-wiki">Personal Wiki</a></h3> <div class="rec-owner">heavenchenggong</div> <div class="rec-desc">个人知识库（LLM Wiki）操作 skill。当用户提到以下意图时触发： - Ingest：处理新内容、更新知识库、"处理IMA新内容"、"处理印象笔记"、"处理raw里的文件"、"帮我ingest" - Query：查 wiki、"w</div> <div class="rec-stats"> <span style="color:#f39c12;">★ 1</span> <span style="color:#5b6abf;">📥 322</span> </div> </div> </div> </div> </div> <script> document.addEventListener('DOMContentLoaded',function(){ document.querySelectorAll('.detail-tab').forEach(function(btn){ btn.addEventListener('click',function(e){ var tab = this.getAttribute('data-tab'); document.querySelectorAll('.detail-tab').forEach(function(b){b.classList.remove('active')}); document.querySelectorAll('.detail-content').forEach(function(c){c.classList.remove('active')}); this.classList.add('active'); var el = document.getElementById('tab-'+tab); if(el) el.classList.add('active'); }); }); }); </script> <div class="footer"> <p>Skill工具集 © 2026</p> </div></body> </html>

Vimeo Locked-Embed Caption Extraction

概述

Vimeo Locked-Embed Caption Extraction

Problem